Metraipolis › archive › dc_5990131bb434

Concept Acupuncturist: An Efficient and Precise Concept Erasure Method for Diffusion Models via Sparse Autoencoders — arXiv:2610.32059

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.32059
License (as archived)
CC0-1.0
Archived
2026-10-26T19:51:38.794Z
SHA-256 of the archived text
903b0e9486b1689af7fda5021da2502e0df79b35633154c997f2b485d19d338f
Stable URL of this page
https://metraipolis.com/documento/dc_5990131bb434

How to cite

Concept Acupuncturist: An Efficient and Precise Concept Erasure Method for Diffusion Models via Sparse Autoencoders — arXiv:2610.32059 — arXiv (en), https://arxiv.org/abs/2610.32059 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5990131bb434 (Metraipolis archive dc_5990131bb434, sha256 903b0e9486b1689af7fda5021da2502e0df79b35633154c997f2b485d19d338f)
BibTeX
@misc{dc5990131bb434,
  title = {Concept Acupuncturist: An Efficient and Precise Concept Erasure Method for Diffusion Models via Sparse Autoencoders — arXiv:2610.32059},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.32059},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5990131bb434; sha256 903b0e9486b1689af7fda5021da2502e0df79b35633154c997f2b485d19d338f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.32059. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Diffusion-based image generation models produce high-quality images but may generate harmful or policy-violating content, motivating machine unlearning (MU) to selectively remove target concepts while preserving general capabilities. However, existing MU approaches often struggle with the distributed and entangled nature of concept representations, leading to inefficient optimization and imprecise edits when applied through large-scale fine-tuning over full diffusion trajectories. To bridge this gap, we propose Concept Acupuncturist (ConceptACP), a novel MU framework that integrates interpretability-guided analysis with inference-free fine-tuning for efficient and precise concept removal. By establishing mappings between semantic concepts and cross-attention blocks in the U-Net denoiser and localizing the most concept-relevant block for targeted optimization, ConceptACP avoids full diffusion trajectory optimization, thereby improving fine-tuning efficiency. To mitigate concept entanglement, we propose a feature-level correction mechanism based on Sparse Autoencoders (SAEs), where SAE-derived sparse, semantically meaningful feature activations are manipulated to construct targeted correction signals, which are then projected back to the output space of the corresponding block and used as supervision for precise concept removal. Consequently, ConceptACP eliminates the need for external inference-time modules while achieving efficient and fine-grained unlearning. Instantiated on SDXL-Turbo, ConceptACP improves the forgetting success rate by 36.34 percentage points in the standard one-step generation setting on average over the best-performing baseline, while preserving image quality and prompt fidelity, and generalizes effectively across other Stable Diffusion backbones. Our code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.32059 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-26T19:45:33.380Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).